Better Understanding of the
Pinellas County Jail Population
Pinellas County Data Collaborative
Pinellas County, Florida
Policy and Services Research Data Center
Department of Mental Health Law & Policy
Louis de la Parte Florida Mental Health Inst.
University of South Florida
The three important factors driving the need for
higher bed capacity are:
1) the number of inmates is increasing over time
2) the length of stays are increasing over time
3) the number of repeat offenders is increasing over time
4) Other factors for Inmate Population growth is the growth in Pinellas
County and mandatory sentencing laws/Policies.
DEMOGRAPHICS
The proportion distribution by demographics has not changed
significantly over time, which means there is no one demographic
characteristic driving the increase of inmates or length of stays.
Although there are the following findings:
•The largest age group population (18 to 25 Year Olds) is also
shows the highest growth (10% a year)
• Although females are still only a small portion of the inmate
population their number (85%) have increase proportionately faster
than the males (50%)
• 77% of the inmate population reside in Pinellas County, Another
12% reside in the three adjacent counties (Hillsborough, Manatee,
Pasco). The other 11% reside mostly in the other Florida Counties
and in the other U.S. states
Average Inmate Population By Gender over Nine
Year Period
Female,
24%
Male, 76%
Average Inmate Population
By Race over nine year period
(1998-2006)
White, 73.04%
Black, 26.07%
Asian, 0.33%
Unknown,
0.55%
American
Indian, 0.01%
Age Group By Gender Over Time
MALES
4,000
3,500
3,000
2,500
2,000
1,500
1,000
500
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
33
32
42
52
77
82
100
114
142
18to25
1,863
1,812
1,974
2,142
2,504
2,684
2,778
2,995
3,363
26to35
2,402
2,193
2,217
2,266
2,258
2,359
2,410
2,459
2,630
36to45
2,127
2,025
2,029
2,084
2,181
2,201
2,253
2,270
2,428
46to64
648
634
725
771
859
942
1,002
1,081
1,129
65to88
46
36
44
42
29
47
61
38
66
89+
0
0
3
0
0
0
0
0
3
Unknown
6
5
2
3
6
4
0
4
28
<=17
<=17
65to88
18to25
89+
26to35
Unknown
36to45
46to64
Age Group By Gender Over Time
FEMALE
12,000
10,000
8,000
6,000
4,000
2,000
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
201
256
306
423
537
627
712
829
967
18to25
6,463
6,561
7,004
7,679
8,151
8,703
9,240
9,669
10,434
26to35
7,767
7,008
6,821
6,922
6,710
6,802
6,988
6,798
7,108
36to45
6,667
6,370
6,451
6,609
6,570
6,397
6,486
6,530
6,422
46to64
3,061
2,979
3,124
3,223
3,357
3,421
3,597
3,694
3,742
65to88
277
256
234
237
253
218
262
228
250
0
0
1
0
2
0
1
0
3
16
11
8
8
4
4
5
10
42
<=17
89+
Unknown
<=17
18to25
26to35
36to45
46to64
65to88
89+
Unknown
County / Non-County Residents
for Inmates Over Time
Distribution of Arrests by Year
County / Non-County Residents
OTHER FLORIDA
COUNTIES
6%
OTHER STATES
3%
NON USA
0%
UNKNOWN
2%
50,000
THREE ADJACENT
COUNTIES
11%
40,000
30,000
20,000
10,000
0
Pinellas
Not Pinellas
PINELLAS
78%
1999
2000
2001
2002
2003
2004
2005
2006
39,233 40,094 42,308 45,919 46,654 46,050 44,200 23,855
7,555
7,328
7,627
8,366
8,710
9,012
8,278
4,254
Over time County Residents made up 77% of the Inmate population
Pinellas and the three surrounding counties made up 89% of the Inmate Population
Across the State of Florida
OTHER
Other is made up the
other 47 Florida Counties
SEMINOLE
ALACHUA
BREVARD
VOLUSIA
DUVAL
MARION
LEE
PALM BEACH
CITRUS
BROWARD
MIAMI-DADE
SARASOTA
ORANGE
HERNANDO
POLK
MANATEE
PASCO
HILLSBOROUGH
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
Across the USA
ARKANSAS
MISSISSIPPI
The only state not showing up is New Hampshire
WISCONSIN
ARIZONA
These state make up 89% of those USA residents
who were arrested in Pinellas County.
LOUISIANA
MISSOURI
SOUTH CAROLINA
All other states were less than 1%
MARYLAND
KENTUCKY
CALIFORNIA
ALABAMA
VIRGINIA
INDIAN
TENNESSEE
NORTH CAROLINA
PENNSYLVANIA
TEXAS
ILLINOIS
MICHIGAN
OHIO
NEW YORK
GEORGIA
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
Average Number of
Inmates Per Day
Average Inmates per Day
Overtime by Gender
1998-2006
3500
3302
3000
2500
2761
2457
2202
2202
2563
2922
2555
2260
2000
Male
Female
1500
1000
500
362
373
410
436
513
1998
1999
2000
2001
2002
515
2003
562
604
667
2005
2006
0
2004
Average Inmates per Day
Overtime by Age Group
1998-2006
1600
1998
1400
114%
1200
1999
16%
27%
2000
1000
2001
84%
58%
800
600
2002
2003
352%
2004
400
0%
200
0%
2005
2006
O
W
N
UN
KN
89
+
65
to
88
46
to
64
36
to
45
26
to
35
18
to
25
<=
17
0
Non-Demographic Indicators
Number of Charges
•The mean number of charges is 1.2 and is consistent overtime, 85% to
87% of the inmate population receive 1 to 2 charges.
• What has changed overtime is the maximum number of charges has
increased from 15 to 99.
•It is the exception rather than the norm when a person received over 4
charges when arrested.
How big is the problem of repeat
offenders?
Repeat , 45%
No Repeat, 55%
• Males (47%) are more likely to be a repeat offender than females (39%)
• African American (57%) are more likely to be a repeat offender than other groups (12%-42%)
• African American Males whose age is <= 17 at their first arrests (72%) are the most likely to be
repeat offenders.
• The younger you are at your first arrest (63%) the more likely you are to be a repeat offender then
other age groups (11%-49%)
How big is the problem of repeat
offenders?
Individuals
Arrests
Single arrest, 23%
Multi-arrests,
44%
Single arrest,
56%
Multi-arrests, 77%
Less than half of the individuals (44%) account for up to 77% of the arrests.
Approximately 15% offenders are arrested again the following year.
Note: It is necessary to look over multiple years to identify a repeat offender
Demographics - Repeat Offender
All
Sex
Race
County Group
Age Group
Male (79%)
Female (21%)
American Indian (<1%)
Asian (<1%)
Black (30%)
White (69%)
Unknown (<1%)
Non USA (<1%)
Other FL County (3%)
Other States (1%)
Pinellas (87%)
3 Adj. Counties (7%)
Unknown (<1%)
<= 17(<1%)
18 to 25 (28%)
26 to 35 (28%)
36 to 45 (28%)
46 to 64 (14%)
65 to 88 (<1%)
89+ (<1%)
Unknown (<1%)
Nbr Arrests
6
6
5
7
4
6
5
5
3
4
3
6
4
5
4
6
6
6
5
4
3
2
Length of stay
3 days
3 days
2 days
2 days
3 days
7 days
3 days
2 days
143 days
7 days
8 days
3 days
4 days
12 days
6 days
2 days
3 days
4 days
4 days
3 days
2 days
8 days
Median Number of days to next arrest
Days, 206
Days, 134
Days, 105
Days, 91
Days, 80
Days, 2
1st
Days, 2
2nd
Days, 5
3rd
Days, 7
4th
Days, 8
5th
Days, 71
Days, 64
Days, 9
Days, 9
6th
7th
Days, 62
Days, 10
8th
Repeat offenders show to have a shorter time between release from jail and their next arrests with each
additional. For example, at their first arrest, they are incarcerated two days and the median days before their
next arrest is 206 days (6-7 months). They repeat this pattern then number of median days before their next
arrests decreases, until they are spending more an more days in jail when arrested and less and less days
out of jail before being re-arrested. For the 7th arrests the median days incarcerated was 9 and then the
median number of day out of jail before being re-arrested was 64 days (2 months).
Number of Arrests
•The breakdown of number of arrests over a nine year period is as
follows:
55% have only one arrest
32 % have up to four arrests
13% have up to 5 arrests
5% have up to 7 arrests
4% have up to 13 arrests
And 1% have up to 85 arrests
•Males on average have 2.5 number of arrests while females have 2.2
•African Americans are more likely to have more arrests, 3.1
•<= 17 year olds are more likely to have more arrests, 3.4
Severe Mental Health Diagnosis and Substance Abuse Diagnosis
The percentage of those found to have a severe mental health diagnosis and/or substance
abuse diagnosis ranged from 5% to 9% over time. It is important to note here that the
identification of any diagnosis was done through matching across the Medicaid System and
the IDS System (State mental health and substance abuse data system). These reported
numbers are expected to be an underestimate as this process does not allow for identification
of any individual who does not interact with either of these systems.
% of population with identified in IDS or Medicaid with a Substance Abuse or
Mental Health Diagnosis Over time
5.00%
4.50%
4.00%
3.50%
Substance
Abuse
3.00%
Mental
Health
2.50%
2.00%
Dual
Diag
1.50%
1.00%
0.50%
0.00%
1998
1999
2000
2001
2002
2003
2004
2005
Figure 9.
2006
Median number of arrests of population with identified in IDS or Medicaid with a
Substance Abuse or Mental Health Diagnosis overtime
7
6
5
Substance
Abuse
4
Mental
Health
3
Dual
Diag
None
2
1
0
1998
1999
2000
2001
2002
2003
Figure 10.
2004
2005
2006
Parole or Conditional Release Violation
 Of the total inmate population it was found that 14% had at least one parole or
conditional release violation.
 Males (15%) are more likely to violate parole or conditional releases then females (13%)
 African Americans (19%)were more likely to violate parole or conditional release
 There are four age groups who were more likely to violate parole or conditional
releases:
o <= 17 years of age
18%
o 18 to 25 years of age
16%
o 26 to 35 years of age
15%
o 36 to 45 years of age
15%
Felony and Misdemeanor Charges








64% of the inmate population have only misdemeanor charges
18% of the inmate population have only Felony charges
18% of the inmate population have both felony and misdemeanor charges
< 1% have neither a felony or misdemeanor charge
35% of inmates have had at least one felony charge
Males (37%) are more likely to have a felony charge than females (30%)
African Americans (52%)are more likely to have a felony charge
Three age groups who are more likely to have a felony charge are:
o <= 17 years olds
38%
o 18 to 25 year olds
37%
o 26 to 35 year olds
37%
Using the Arrest Statutes Literal six
types of Crime were created.
Drug
Sex
Property
Moving
Violent
Other
How they were defined and created can be found in Appendix B in the Report





Drug
41% of the inmate population has at least one crime type of drug
Males (43%) are more likely than females to have a crime type of drug
American Indian (43%) are more likely to have a crime type of drug
Whites (43%) are more likely to have a crime type of drug
Ages 36 to 45 (44%), and ages 46 to 64 (43%) are more likely to have a crime type of
drug




Moving
22% of the inmate population had at least one crime type of moving
Males (24%) were more likely than females to have a crime type of moving
African American (28%) were more likely to have a crime type of moving
Ages <= 17 (55%), and ages 18 t o25 (28%), and ages 26 to 35 (24%), ages 89+ (36%)
were more likely to have a crime type of moving





Property
29% of the inmate population had at least one crime type of property
Females (35%) were more likely than males to have a crime type of property
African American (34%) were more likely to have a crime type of property
American Indians (43%) were more likely to have a crime type of property
Ages <= 17 and ages 18 to 25 (33%) were more likely to have a crime type of property
Sex
Only 4% of the inmate population had at least one crime type of sex
Even though females (3%) had a slightly lower rate of this type of crime, there is a
difference of the type of sex crime by gender.
Asian (5%) were more likely to have a crime type of sex
Ages 65 to 88 (7%) more likely to have a crime type of sex




Violent
26% of the inmate population has at least one crime type of Violent
Males (27%) were more likely than females (24%) to have a crime type of violent
Asian were (29%) more likely to have a crime type of violent
African American were (31%) more likely to have a crime type of violent
Ages <= 17, ages 26 to 35, and ages 36 to 45 were (27%) more likely to have a crime
type of violent








Other
22% of the inmate population has at least one crime type of Other
African American (28%) were more likely to have a crime type of other
Ages <= 17 (26%) and ages 26 to 25 (23%) were more likely to have a crime type of
other
Violent Weapon Involved




Less than 2% of the inmate population showed to have a violent weapon during the
crime arrest
Males (2%) were more likely than females (0.73%) to have a violent weapon during the
crime arrest
African American (3%) were more likely to have a violent weapon during the crime
arrest
Ages <= 17 (4%) were more likely to have a violent weapon during the crime arrest
Minor Involved




Only 2% of the inmate population had a crime arrest involving a minor
Females (2.4%) were more likely to have a crime arrest involving a minor
Asian (3%) and American Indian (10%) were more likely to have a crime arrest involving
a minor
Ages 26 to 35 (2.18%),and ages 36 to 45 (2.18%) and, ages 65 to 88 (2.19%) were
slightly more likely to have a crime arrest involving a minor
Elder and/or Disabled Person Involved




0.24% of the inmate population had a crime arrest involving an elder/disabled person
Females (0.33%) are more likely than males (0.20%) to have had a crime arrest
involving an elder/disabled person
Whites (0.27%) are more likely to have had a crime arrest involving an elder/disabled
person
Ages 46 to 64 (0.65%), and 65 to 88 (1.73%) are more likely to have had a crime arrest
involving an elder/disabled person
Other Systems
Interaction
Emergency Medical System Interaction




12% of the inmate population had interaction with EMS
Females (16%) are more likely than males (11%) to interact with EMS
African American (14%) are more likely to interact with EMS
Ages <= 17 (16%) and ages 36 to 45 (14%), and ages 46 to 64 (17%), and ages 65 to
88 (22%) are more likely to interact with EMS
Department of Social Service System Interaction (NOTE: 1998 – 2003 data only)





9% of the inmate population had interaction with DSS
Females (13%) were more likely than males (8%) to have had interaction with DSS
African American (15%) are more likely to have had interaction with DSS
Ages 36 to 45 (13%), and ages 64 to 64 (12%) are more likely to have had interaction
with DSS
The breakdown of those in the CJIS system also interacting with DSS (16,170) at least
once from (1998 through 2003) by the three types of clients DSS has is as follows:
o Client
9,861
61%
o Depend
3
<1%
o Homeless
2,033
13%
o Depend & Client
129
<1%
o Homeless & Client
4,128
26%
o All 3 types
16
<1%
Mental Health / Substance Abuse Data System Interaction




5.5% of the inmate population had interaction with IDS
Females (8%) were more likely than males (5%) to have had interaction with IDS
Whites (5.57%) are slightly more likely to have had interaction with IDS
Ages <= 17 (6.5%) and ages 26 to 35 (5.57%), and ages 36 to 45 (7%) are slightly more
likely to have had interaction with IDS
Medicaid Data System Interaction




5.5% of the inmate population had at least one interaction with the Medicaid System
Females (7%) are more likely than males (5%) to have had interaction with Medicaid
African American (7%) are more likely to have had interaction with Medicaid
Ages 36 to 45 (7%), and ages 46 to 64 (11%), and ages 65 to 88 (20%), and ages 89+
(9%) are more likely to have had interaction with Medicaid
Baker Act System

Approximately 1% to 3% of the inmate population in any of the nine years has
interacted with the Baker Act System. This is the Florida Involuntary 72-hours
commitment process where individuals are placed in an agency to have a mental health
assessment of danger to themselves or others.
Also note that the custody status of mental health commitment of inmates has increased
from .08% to 0.20% over the last nine years.

Note: There are not identifiable information other than date of birth and gender in this file to
link across systems by individuals. Probabilistic Population Estimation (PPE) was used to
examine the overlap between the Baker Act System to the CJIS system as well as using
those inmates who could be identified using the Medicaid and Statewide Mental Health and
Substance Abuse Systems (9,514 inmates identified), 3,330 inmates in the CJIS/Jail
System were identified in the Baker Act System (35%).
Figure 11. Inmates Arrested who also have receiving a Baker Act Initiation at some
point in time over the nine years
Nbr Inmates
3,000
2,572
2,500
2,256
1,954
2,000
1,815
1,695
1,500
1,438
1,470
2000
2001
Nbr Inmates
1,000
643
500
0
1999
2002
2003
2004
2005
2006
Average Length of Stay
Overtime
Length of Stay
Table 7. Overall
Number of
Arrests
All
Sex
Race
Age
Group
Male
Female
American Indian
Asian
Black
White
Unknown
<= 17
18 to 25
26 to 35
36 to 45
46 to 64
65 to 88
89+
Unknown
Number of Arrests
Total
Repeat
Population
Offenders
4
6
4
6
3
5
1
4
2
4
5
6
3
5
1
3
3
4
4
4
3
2
1
1
4
5
5
5
5
4
2
2
Length of Stay
Total
Repeat
Population
Offenders
2 days
3 days
3 days
3 days
2 days
3 days
2 days
2 days
2 days
3 days
4 days
6 days
2 days
3 days
2 days
4 days
2 days
2 days
2 days
3 days
2 days
2 days
1 days
2 days
3 days
2 days
3 days
4 days
days
3 days
1.5 days
3 days
Table 9. Gender
There was not a significant difference on the median length of stay by gender, even though
males do show a slightly longer length of stay than females.
1998
1999
2000
2001
2002
2003
2004
2005
2006
Female
2 days
2 days
2 days
2 days
2 days
2 days
2 days
2 days
2 days
Male
2 days
2 days
2 days
2 days
2 days
3 days
3 days
3 days
2 days
Figure 12. Race
Two things are apparent when looking at length of stay by race. One, that African Americans
length of stay is significantly longer than other races and two, this holds true over time.
6
1998
5
1999
2000
4
2001
3
2002
2003
2
2004
2005
1
2006
0
American
Indian
Asian
Black
White
Unknown
Figure 13. Age Group
It is important to note that there are very low number in three age categories (<=17, 89+,
UNKNOWN). Any extreme changes overtime in these groups are influenced not by a group
pattern but usually one individual. During the first five years (1998-2002) it shows that <= 17
age category stayed significantly longer than other groups, but more importantly in the recent
four years the median length of stay for those <= 17 year of age is significantly less and below
the median of other age categories.
25
1998
20
1999
2000
15
2001
2002
10
2003
2004
2005
5
2006
O
W
N
UN
KN
89
+
65
to
88
46
to
64
36
to
45
26
to
35
18
to
25
<=
17
0
Table 9. The number of charges
The median length of stay does increase as a function of increases in the number of
charges. Note that 85%-87% receive only 1 to 2 charges and 99% of individuals never
receive more than 5 charges during one arrest. In 2006, if an individual was arrested and
had four to five charges, the median length of stay was approximately 17 to 20 days.
Charge Counts
1998
1999
2000
2001
2002
2003
2004
2005
2006
1
2
2
2
2
2
2
2
2
2
2
9
9
9
11
12
10
10
8
9
3
11
16
13
22
20
14
12
12
13
4
5
22
23
14
14
26
17
18
20
20
6
20
25
25
38
39
20
27
22
17
7
22
22
22
28
35
22
21
17
22
8
46
21
54
16
50
62
44
24
61
9
35
48
16
9
23
14
47
42
6
16
15
22
54
36
37
102
56
14
10
> 10
49
35
28
38
29
23
52
40
22
53
12
44
29
39
43
12
63
23
The number of arrests
The median length of stay does increase with the number of arrests, but is not a strong factor
that drives length of stay.
 An individual with one arrests the median length of stay is 2 days
 An individual with 4 arrests the median length of stay is 3 days.
 An individual with 5 arrests the median length of stay is 4 days
 An individual with 7 arrests the median length of stay is 5 days
 An individual with 13 arrests the median length of stay is 8 days
Figure 17. Receiving a Parole or Conditional Release Violation

Having a parole or conditional release violation is highly correlated to the number of
days incarcerated.
14
12
10
Parole and/or Cond. Release
Violation
8
No Parole and/or Cond.
Release
Violation
6
4
2
0
1998 1999 2000 2001 2002 2003 2004 2005 2006
Failure to Appear

The was no significant difference in length of stay due and Failure to Appear, although
analysis did show those with Failure to Appear spent less time incarcerated than the
average median stay. This maybe due to their type of crime.
Alcohol Involved in Arrest

There was no significant difference in the length of stay and alcohol involvement during
the arrest.
Drugs Involved in Arrest

Drugs being involved in the arrests show to significantly increase the length of stay. The
median number of days incarcerated for arrests where drugs were found to be involved
is 6 days.
Felony Charge

Having a felony charge at the time of arrest significantly correlated with an increase the
length of stay. The median number of days incarcerated for arrests where there was at
least one felony charge is 13 days.
Table 11. Crime Type

Length of stay did have a significant increase not only for the felony by crime type. The
highest length of stays being for sex and violent crime types, then drug crimes and lastly
moving crimes.
1998
1999
2000
2001
2002
2003
2004
2005
2006
Drug (F)
11 days
12 days
10 days
14 days
16 days
13 days
14 days
13 days
12 days
Moving (F)
3 days
3 days
3 days
5 days
4 days
3 days
3 days
2 days
2 days
Other (F)
13 days
14 days
14 days
14 days
11 days
10 days
10 days
8 days
6 days
Property (F)
14 days
15 days
13 days
17 days
20 days
17 days
15 days
10 days
9 days
Sex (F)
22 days
28 days
22 days
29 days
31 days
24 days
23 days
32 days
57 days
Violent (F)
16 days
16 days
15 days
18 days
22 days
17 days
16 days
15 days
12 days
Violent Weapon Involved

Having a violent weapon at the time of arrest show significantly increase the length of
stay. The median number of days incarcerated for arrests where there was a violent
weapon at the time of arrest is 12 days.
Crimes involving Minors, Elders, and/or Disabled persons

Crimes involving Minors, Elders and/or Disabled persons did not have an influencing
factor to the length of stay.
Interaction with Emergency Medical Services System

Those who interact with EMS have a median total days of 11 compared to the median
total days of 3 for those who do not show having interacted with EMS.
Interaction with Dept. of Social Services System

Those who interact with DSS have a median total days of 34 compared to the median
total days of 3 for those who do not show having interacted with DSS.
Interaction with Medicaid System

Those who interact with Medicaid have a median total days of 10 compared to the
median total days of 4 for those who do not show having interacted with Medicaid.
Interaction with State Mental Health and Substance Abuse System

Those who interact with IDS have a median total days of 27 compared to the median
total days of 3 for those who do not show having interacted with IDS
Figure 18. Bond Levels

There was a relationship with bond level and the length of stay, but it also has a
relationship to the type of charge (felony / misdemeanor). And the data also showed the
there were always those who had a high bond that were in the median length of stay.
Being able to bond out has a lot to do with the economic status of the individual and
there is a concern to link length of stay to bond levels until further analysis is done.
Median Bond Amount
$25,000
$30,000
$21,125
$10,000
$25,000
$20,000
$15,000
$10,000
$5,000 $5,000
$2,013 $2,500 $2,500 $5,000
$2,013 $2,500
$5,000
$0
ee
k
2
w
ee
ks
3
be
w
ee
tw
ee
ks
be n 1 1 m
tw
an
on
ee
d
2 th
be n 2
m
tw
on
an
ee
th
d
s
3
be n 3
m
tw
on
an
ee
th
d
s
4
be n 4
m
tw
an
on
ee
d
th
5
s
be n 5
m
tw
on
an
ee
th
d
s
6
be n 6
m
tw
on
an
ee
th
d
s
7
be n 7
m
tw
an
on
e
d
th
be en
8
s
m
8
tw
o
a
ee
nt
nd
be
hs
n
9
9
tw
m
a
ee
nd
on
th
be n 1
10
s
0
tw
m
an
ee
on
d
n
th
11
11
s
m
an
o
nt
d
hs
12
O
m
ve
r 1 onth
2
s
m
on
th
s
1
w
s
$500
da
y
3
th
an
s
Le
s
$7,500$7,500
$5,000
$5,000 $5,000
Figure 19. Custody Status
Since 2004 the number of released and released on their own recognizance has gone down
corresponding to the number of out of bond and maximum security going up.
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
1998
1999
2000
2001
2002
2003
2004
2005
BOND
RELEASED
RELEASED ON OWN RECOGNIZANCE
MAXIMUM SECURITY CENTER
TO CORRECTIONAL FACILTY OR PRISON FACILITY
JAIL: CENTRAL DIVISION
MEDIUM SECURITY CENTER
FEMALE MIN. SEC. CENTER
FEMALE MAX. SEC. CENTER
SPECIAL SECURITY:
OUT OF THE COUNTY OR OUT TO HOSPITAL
MALE MIN. SEX.CENTER
OLD BOOT CAMPE ANNEX BLDG
TEMP. HOLDING CELL FOR TRANSPORT
ELECTRIC MONITORING - SUBJECT NOT IN JAIL
OUT ON FURLOUGH
OUT TO HOSPITAL
2006
Table 12. Custody Status (continued)

There is some difference of length of stay associated with custody status (where the
inmate is housed), but this more has to do with the gender and level of crime than
directly to length of stay.
1998
1999
2000
2001
2002
2003
2004
2005
2006
1998
1999
2000
2001
2002
2003
2004
2005
2006
Jail: Central Division
JACD
7 days
25 days
10 days
16 days
9 days
7 days
21 days
Female Max Sec Ctr
JAFC
20 days
16 days
7 days
11 days
6 days
21 days
Male Min Sec Ctr
JAMN
1 days
16 days
2 days
4 days
3 days
5 days
Female Min Sec Ctr
JAFN
2 days
5 days
18 days
99 days
4 days
6 days
4 days
3 days
13 days
Male Med Sec Ctr
JAMS
2 days
3 days
34 days
5 days
4 days
18 days
4 days
13 days
Male Max Sec Ctr
JAMX
71 days
14 days
77 days
61 days
40 days
42 days
16 days
34 days
73 days
Old Boot Camp
Temp Holding Cell
Annex Bldg - JAND
JASO
8 days
9 days
2 days
3 days
3 days
35 days
3 days
37 days
15 days
44 days
ACTIVE CASES IN JAIL IN and OUT OF JAIL
Cases were identified as active included all cases except those with the court disposition
status as one of the following codes: BOND, FURL, HOSP, JAEM, JAGW, JAOT, METN,
OREC, PROB, PRST, RLSD, and VOID. In 2006, only 3% of the cases showed to have a
custody status where they were incarcerated (JACD, JAFC, JAFN, JAMN, JAMS, JAMX,
JAND, JASO).
Figure 20. THOSE OUT OF JAIL, OUT ON WHAT STATUS
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
1998
1999
2000
2001
2002
2003
2004
2005
2006
BOND
RELEASED
RELEASED ON OWN RECOGNIZANCE
TO CORRECTIONAL FACILTY OR PRISON FACILITY
SPECIAL SECURITY:
OUT OF THE COUNTY OR OUT TO HOSPITAL
MENTAL HEALTH COMMITMENT
ELECTRIC MONITORING - SUBJECT NOT IN JAIL
One way to look at jail bed usage is to look at the number of inmates as consumers of jail
bed days. Some consumers use more jail bed days than others. Below is what is called a
Lorenz Curve, which is a graphical representation of the cumulative distribution
function of a probability distribution; it is a graph showing the proportion of the
distribution assumed by the bottom y% of the values. In this case, this graph is the used
to represent the jail bed usage of inmates. A perfectly equal usage distribution would be
one in which every inmate uses the same number of jail bed days (Blue diagonal line).
The actual distribution of jail bed days by inmates is a line of inequality (Pink curved
line), which shows that 65% of the population use only 3% of the jail bed days, another
30% of the population use 51% of the jail bed days and the last 5% of the inmate
population use 46% of the jail bed days.
65% of Inmate
Population / 3%
of Jail Bed
Days
1)
Three groups
Low Bed Users (LBU),
2)
2) High Bed Users (HBU)
1)
and 3) Greatest Bed
Users (GBU)
5% of Inmate
Population /
46% of Jail Bed
Days
Table 13. & Table 14. Demographics (Gender, Race, and Age Categories)
None of the demographics categories (Gender, Race, Age Group) showed any specific pattern
across the three groups by demographics (% within each of the three groups). Examining the
distribution across each of the demographic categories, you can see that males, African
Americans, and those <= 17 years of age at first arrest show to be more likely in the Greatest
Bed Users than females, other races, and other age groups.
%
Male
Female
American Indian
Asian
African American
White
Unknown
<= 17 Years of age
18 to 25 Years of age
26 to 35 Years of age
36 to 45 Years of age
36 to 64 Years of age
65 to 88 Years of age
89 + Years of age
Unknown
% within each of Three Group
LBU
HBU
GBU
70%
81%
86%
30%
19%
14%
<1%
<1%
1%
<1%
<1%
16%
29%
48%
82%
71%
52%
2%
1%
<1%
1%
2%
5%
29%
29%
32%
27%
28%
29%
25%
28%
25%
17%
13%
8%
2%
1%
<1%
<1%
<1%
<1%
-
% of each Demo Category
LBU
HBU
GBU
61%
33%
6%
75%
23%
2%
90%
10%
75%
21%
4%
48%
41%
11%
69%
28%
3%
69%
31%
<1%
44%
38%
18%
64%
30%
6%
63%
32%
5%
62%
33%
5%
71%
26%
3%
83%
15%
2%
100%
82%
18%
-
Looking at the median length of stay for each of the three groups by demographics shows the
extreme difference between the LBU from either the HBU and the GBU.
LBU
2 days
HBU
72 days
GBU
482 days
Gender
LBU
2 days
2 days
HBU
62 days
74 days
GBU
222 days
254 days
RACE
LBU
1 days
2 days
2 days
2 days
2 days
HBU
128 days
78 days
91 days
65 days
44 days
GBU
496 days
501 days
465 days
457 days
AGE GROUP
<=17
18to25
26to35
36to45
46to64
65to88
89+
UNKNOWN
LBU
1
2
2
2
2
2
2
2
HBU
126
77
71
71
59
44.5
43.5
GBU
501
477
484
482
483
458
-
ALL
Females
Males
American Indian
Asian
African American
White
Unknown
Table 15. Other
Non-demographic Indicators
LBU
The non-demographic indicators that seem to
identify difference between the three groups are
Repeat offender, level of crime
(Felony/Misdemeanor), Number of arrests, a
violation of parole or conditional release.
Other factors were DSS interaction, which
needs further investigation to understand;
number of years in the CJIS system, which really
can be explained that the more years in the CJIS
system, the more arrests and days incarcerated;
and the type of crime also showed a consistent
increase across groups.
Number of Arrests
Age at First Arrest
Number of Years in CJIS System
Parole or Conditional Release Violation
Failure to Appear
Felony Only
Misdemeanor Only
Both Felony and Misdemeanor
None
Substance Abuse Diag only
Severe Mental Health Diag Only
Dual Diagnosis
No Diagnosis found
EMS Interaction
IDS Interaction
Medicaid Interaction
DSS Interaction
Elder/Disabled Person Involved
Minor Involved
Violent_weapon at arrest
Drug Crime
Property Crime
Sex Crime
Violent Crime
Moving Crime
Other Crime
Drug Involved
Alcohol Involved
Repeat Offender
1
34
1
7%
11%
13%
81%
5%
<1%
2%
2%
<1%
96%
10%
4%
5%
6%
<1%
2%
1%
36%
22%
3%
22%
18%
17%
9%
22%
24%
HBU
GBU
4
33
2
28%
13%
25%
38%
36%
1%
4%
3%
1%
92%
16%
8%
7%
14%
<1%
2%
2%
50%
39%
5%
31%
28%
27%
18%
13%
80%
8
31
4
29%
12%
18%
7%
74%
<1%
5%
6%
2%
87%
21%
13%
8%
24%
<1%
2%
4%
65%
58%
13%
54%
35%
46%
22%
7%
94%
Interactions with more than two systems
The majority of the CJIS population does not interact with other systems (76%). Of those who
do interact with other systems, 22% interact with 1 or 2 other systems. There is approximately
2 % of the population, who interact with 3 to all 4 systems. (Table 16.)
CJIS Only
CJIS & EMS Only
CJIS & DSS Only
CJIS & AHCA Only
CJIS & IDS Only
CJIS & EMS & DSS
CJIS & EMS & IDS
CJIS & EMS & AHCA
CJIS & DSS & AHCA
CJIS & DSS & IDS
CJIS & EMS & DSS & IDS
CJIS & EMS & DSS & AHCA
CJIS & IDS & Medicaid
CJIS & EMS & IDS & ACHA
CJIS & EMS & DSS & IDS & ACHA
CJIS & DSS & IDS & ACHA
133124
12932
9103
3949
3934
2560
1888
1644
1215
909
874
795
698
571
385
329
76%
7%
5%
3%
3%
2%
1%
1%
<1%
<1%
<1%
<1%
<1%
<1%
<1%
<1%
Length of stay over 365 days
1% of the inmate population length of stay is over one year, and median of 479 days. These individuals
use up on average 10% of the jail day beds each year.
Mental Health / Substance Abuse / Dual / and NO Diagnosis
Since there is an interest specifically in mental health and substance abuse and interaction
with the CJIS system, further analysis were done to help understand this population, including
the breakdown by diagnosis, the specific types of diagnosis, and the interactions with DSS and
the EMS systems.
There were 9,596 individuals where were identified in the CJIS system to have either a severe
mental illness diagnosis or a substance abuse diagnosis or both. The breakdown is as follows:




Severe Mental Health Diagnosis:
Substance Abuse Diagnosis:
Dual Diagnosis:
None identified:
3,927
4,242
1,427
165,314
/
/
/
/
2.25%
2.43%
< 1%
94.51%
1,095
1,554
523
35,583
/
/
/
/
1.35%
4.01%
1.35%
91.82%
In 2006 the breakdown was as follows:




Severe Mental Health Diagnosis:
Substance Abuse Diagnosis:
Dual Diagnosis:
None identified:
Of those with a Severe Mental Health Diagnosis, in 2006, the breakdown of diagnosis is as
follows:
 Schizophrenic Disorders
22%
 Episodic Mood Disorders
75%
 Delusional Disorders
<1%
 Other Non-organic Disorders
4%
Of those with a substance abuse diagnosis, in 2006, the breakdown of diagnosis is as follows:



Non-Dependence Drug Use
Alcohol Dependence
Drug Dependence
35%
27%
44%
Geographic
Information Systems
(GIS) Mapping of Inmate
Population using residential zip
codes
The GIS piece of this paper was
done by Luis Perez, a PhD
student in Education at USF, as
part of his course work
requirements.
Overall: As stated in the section
examining residency status of
inmate the majority of the inmate
population reside in Pinellas
County, and where there is
increased residential population
density in Pinellas County there
is also an increase in the density
of residency of the inmate
population.
In the three surrounding counties
there are pockets where 1 to 10
of the Pinellas inmate population
resides.
BY GENDER: Even remembering that Males are the majority of the Pinellas
CJIS inmate population, the zip codes within Pinellas county show to be similar
between Males and Females when mapped. However Males within the three
adjacent counties are coming from a wider spread geographic area (more zip
code areas) than females.
BY AGE GROUP: Of all the eight age groups, the youngest (<=17), and oldest (65 to 88)
age groups show to reside mainly within the county of Pinellas. This is important
information, especially for the youngest age group, because it tells us that if any programs
focusing on decreasing the number of <= 17 year olds from interacting with the CJIS
system, should work within Pinellas county. The study already showed when the younger
you are when you interact with CJIS, the more likely that you will be a repeat offender and
potentially become a GBU. The other age groups seem to increase and spread out more
across the three adjacent counties as the age increase.
Recommendations:






Examine closer the types of interactions CJIS population have with the other systems,
looking for patterns of demographics, services received and over time. There maybe
patterns from the order in which an individual or group of individuals flow through and
between systems
Examine closer when a LBU moves to a HBU and/or GBU and potential indicators to
look at when identifying these individuals. The largest inmate population is the LBU.
Most HBU and/or GBU inmates got put into these 2 categories overtime, types of
crimes, and number of arrests.
Incorporating case studies and in-house studies to answer the questions that the data
housed through the Pinellas Data Collaborative could not answer.
o Review of notices to appear over time - Unknown how to identify these
individuals
o Review of housing and services for inmate upon release - Data not collected by
the Data Collaborative
o Review of programs/education for inmate during incarceration - Data not
collected by the Data Collaborative
o Correlation between CJIS/jail and homeless - Data not yet collected by the Data
Collaborative
A Sub-study to examine patterns of those who have volunteered for drug court
A sub-study to look at those inmates who can also be found in the Dept of Juvenile
Justice to see if any indicators can be found to identify youth who are more likely to
enter into the CJIS jail system over time and programs to prevent this from happening.
A evaluation of those who are HBU, GBU to see if the numbers can be decreased,
decrease their length of stay, or divert them to prison system. Also evaluate those who
are LBU and see if the numbers can be decreased, through non-arrest, early release,
diversion to other programs, etc.
Recommendations:


Examine closer the types of interactions CJIS population have with the other systems,
looking for patterns of demographics, services received and over time. There maybe
patterns from the order in which an individual or group of individuals flow through and
between systems
Examine closer when a LBU moves to a HBU and/or GBU and potential indicators to
look at when identifying these individuals. The largest inmate population is the LBU.
Most HBU and/or GBU inmates got put into these 2 categories overtime, types of
crimes, and number of arrests.
Recommendations Continued….




Incorporating case studies and in-house studies to answer the questions that the data
housed through the Pinellas Data Collaborative could not answer.
o Review of notices to appear over time - Unknown how to identify these
individuals
o Review of housing and services for inmate upon release - Data not collected by
the Data Collaborative
o Review of programs/education for inmate during incarceration - Data not
collected by the Data Collaborative
o Correlation between CJIS/jail and homeless - Data not yet collected by the Data
Collaborative
A Sub-study to examine patterns of those who have volunteered for drug court
A sub-study to look at those inmates who can also be found in the Dept of Juvenile
Justice to see if any indicators can be found to identify youth who are more likely to
enter into the CJIS jail system over time and programs to prevent this from happening.
A evaluation of those who are HBU, GBU to see if the numbers can be decreased,
decrease their length of stay, or divert them to prison system. Also evaluate those who
are LBU and see if the numbers can be decreased, through non-arrest, early release,
diversion to other programs, etc.
Analyses within Bed User Group
Factors
Felony
Crime Type Sex
Crime Type Violent
Crime Type Drug
African American
DSS Interaction
Male
Crime Type Moving
group comparision
Odd Ratio
CI for Odd Ratio
Greatest Bed Users vs All others
14.268
( 13.177, 15.450)
Greatest Bed Users vs All others
5.249
( 4.837, 5.697)
Greatest Bed Users vs All others
3.239
( 3.086, 3.401)
Greatest Bed Users vs All others
2.459
( 2.339, 2.339)
Greatest Bed Users vs All others
2.210
( 2.104, 2.320)
Greatest Bed Users vs All others
2.093
( 1.970, 2.223)
Greatest Bed Users vs All others
1.932
( 1.804, 2.068)
Greatest Bed Users vs All others
1.633
( 1.550, 1.719)
p-value
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
Note:
(1) *The interpretation: The Greatest Bed Users are 14.268 times more likely to have at least one felony charge then those in the Low
Bed User group.
(2) The p-value less than 0.05 means there is a significant difference between the two groups.
Analyses within Bed User Group
Factors
Felony
DSS Interaction
IDS Interaction
Male
group comparision
Odd Ratio
CI for Odd Ratio
High Bed Users vs Low Bed Users
6.537
( 6.381, 6.697)
note: omited the Greatest Bed User group
High Bed Users vs Low Bed Users
2.230
( 2.141, 2.323)
note: omited the Greatest Bed User group
High Bed Users vs Low Bed Users
2.048
( 1.946, 2.157)
note: omited the Greatest Bed User group
High Bed Users vs Low Bed Users
2.024
( 1.966, 2.082)
African American High Bed Users vs Low Bed Users
1.629
note: omited the Greatest Bed User group
Failure to Appear High Bed Users vs Low Bed Users
1.512
note: omited the Greatest Bed User group
EMS Interaction High Bed Users vs Low Bed Users
1.434
note: omited the Greatest Bed User group
Drug Involvement High Bed Users vs Low Bed Users
1.391
note: omited the Greatest Bed User group
Medicaid Interaction High Bed Users vs Low Bed Users
1.112
note: omited the Greatest Bed User group
( 1.583,
p-value
<.0001
<.0001
<.0001
<.0001
1.675)
<.0001
( 1.459, 1.566)
<.0001
( 1.384,
1.487)
<.0001
( 1.344, 1.440)
<.0001
( 1.056,
<.0001
1.170)
Note:
(1) *The interpretation: The High Bed Users are 6.537 times more likely to have at least one felony charge then those in either the High
Bed User group or the Low Bed User group.
(2) The p-value less than 0.05 means there is a significant difference between the two groups.
Purposes and Uses of this presentation:
This presentation was generated in response to specific questions posed by
member of the Pinellas Data Collaborative. It was created to inform administrative
policy and program decisions that benefit the citizens of Pinellas County. Before
reusing or citing findings in this report, please contact the Data Collaborative to ensure
accurate understanding of the analyses and interpretation of results. Questions should
be
directed to Diane Haynes at dhaynes@fmhi.usf.edu or 813-974-2056.